Similar model-based methodologies have been successfully implemented in areas facing similar gaps in evidence: for instance

The gap

Evidence profile

Sourced from the future work and conclusions of the source papers, classified as general, spanning 3 journals.

Supporting evidence — 3 representative gaps

• Advancing Pediatric Dose Scaling: Strategies, Modeling Approaches, and Clinical Applications (2026) · Pharmaceuticals · doi

  Model-informed drug development will increasingly guide pediatric dose selection  across drug development phases. Integration of in silico, in vitro, and in vivo data into  detailed PBPK and PopPK models will enable more accurate predictions, particularly for  neonates and infants.  Novel  biomarkers  beyond  SCr,  including  cystatin  C  and  emerging  kidney  injury  markers, may improve renal function estimation in young children. Machine learning ap- proaches are being explored to identify optimal covariate relationships and improve the  predictive performance of PopPK models. Real-world data from EHRs offer opportunities  for  post-marketing  dose-optimization  studies  that  were  previously  impractical.  Digital  twins  represent  an  emerging  evolution  of  model-informed  precision  dosing,  in  which  mechanistic pharmacometric models are combined with individual patient physiological  and clinical data to form a virtual patient that is updated as new data accrue, allowing  simulation of drug exposure, prediction of therapeutic response and toxicity, and adap- tive  dose  optimization  across  the  pediatric  age  spectrum  [121,126].  Recent  reviews  de- scribe an emerging paradigm in which machine learning and artificial intelligence meth- ods are integrated with PBPK and PK/PD modeling to support parameter estimation, vir- tual population generation, and uncertainty quantification [127]. Critical assessments em- phasize, however, that these approaches must be evaluated against established pharma- cometric standards, with explicit attention to interpretability, training-set diversity, and  prospective  validation  before  clinical  adoption  [128].  Trial-design  methodology  is  also  evolving: an accuracy-for-dose-selection framework has been proposed as an alternative  to traditional parameter-precision criteria for justifying pediatric PK study designs, with  the aim of more directly aligning trial designs with the regulatory question of selecting an  appropriate dose [129].  Integration of adaptive dosing approaches with therapeutic drug monitoring (TDM),  Bayesian forecasting, and model-informed precision dosing tools will further refine indi- vidualized treatment in clinical practice. Emerging technologies, including artificial intel- ligence, digital  twins, and learning  healthcare systems, are  likely to complement estab- lished pharmacometric approaches by enabling continuous refinement of dosing recom- mendations as new patient data become available.  Looking  ahead,  MIPD,  already  established  for  narrow-therapeutic-index  agents,  is  expected to broaden across drug classes in pediatric practice as it is integrated with trans- porter and enzyme ontogeny data and emerging real-world data sources, although ran- domized  trials  demonstrating  clinical  benefit  are  still  needed.  Recent  reviews  illustrate  how PBPK modeling can complement MIPD by supplying mechanistic predictions of site- of-infection PK/PD targets, drug–drug interactions, and exposure in special subpopula- tions such as preterm neonates, obese children, and those with renal impairment that pop- ulation models alone may not adequately capture [121].  Important gaps remain. Mechanistic data are still sparse for preterm neonates, trans- porter and enzyme ontogeny, biologics and immunogenicity in young children, pediatric  PD,  and  disease states  such  as  ARC and  therapeutic  hypothermia.  Continued progress  will depend on generating high-quality pediatric clinical data and refining ontogeny and  disease-state physiology in pediatric PBPK platforms, prospectively validating model-in- formed approaches in clinical studies, and demonstrating that improved exposure predic- tion translates into improved clinical outcomes across diverse pediatric populations.  Ultimately,  the  integration  of  therapeutic  drug  monitoring,  Bayesian  forecasting,  model-informed  precision  dosing,  artificial  intelligence,  digital  twins,  and  learning  healthcare systems has the potential to deliver increasingly individualized pediatric ther- apy. The goal is that every child receives a dose informed by the best available mechanis- tic, clinical, and regulatory evidence.  https://doi.org/10.3390/ph19071090    Pharmaceuticals 2026, 19, 1090  27  of  35

  generalfuture work
  Keywords: pediatric drug clinical dose model informed emerging dosing therapeutic across pbpk models learning precision approaches
• DRUG REPURPOSING USING ARTIFICIAL INTELLIGENCE AND NETWORK PHARMACOLOGY FOR NEURODEGENERATIVE DISEASES: A COMPREHENSIVE REVIEW (2026) · Zenodo (CERN European Organization for Nuclear Research) · doi

  7.1 Experimental and clinical validation Although AI and network-based pharmacology have improved repositioning, experimental and clinical validation is essential to confirm the drug's predicted activities on the target and in experiments to prove its effectiveness in the clinic. The use of real-world data and EHRs for clinical trial emulation complements the preclinical results, by evaluating real-world drug safety and efficacy.[5] 7.2 Data transformation and data dimension Multi-omics and clinical data analysis come with limitations, model sharing challenges data of www.ejbps.com │ Vol 13, Issue 9, 2026. │ ISO 9001:2015 Certified Journal │ 19 Sree N. et al. European Journal of Biomedical and Pharmaceutical Sciences transparency challenges, and difficulties. There are two new strategies namely federative learning features and AI explanation, used to make the data more transparent and private.[5] 7.3 Precision medicine Integrating AI-guided multi-omics and network-based pharmacology helps advance precision medicine by identifying patient subpopulations most likely to benefit from particular repositioned drugs, accounting for factors like sex, genetic background, and disease state.[2,3,5] 8. CONCLUSION AI and network pharmacology create a powerful combination to advance drug repositioning strategies for neurodegenerative diseases, aiding in comprehensive, data-driven drug identification and understanding of their mechanisms. Repurposed drugs with potential vary across various classes such as anti-inflammatory agents and antidiabetics, antihypertensives and epigenetic modulating sex-specific particularities and multi-omics adds further precision in therapeutic approaches.

  generalfuture work
  Keywords: clinical drug network pharmacology multi omics precision experimental validation based repositioning real world challenges journal
• Supporting clinical guidelines for opioid conversion to methadone and tapering to prevent withdrawal in critically ill children using physiology based pharmacokinetic modeling and simulation (2026) · PLoS ONE · doi

  Similar model-based methodologies have been successfully implemented in areas facing similar gaps in evidence: for instance, in optimizing gentamicin dosing in neonates and infants [8], determining age-specific dexamethasone doses to prevent post-extubation stridor in children [51], and supporting drug use in pregnancy where clinical data are limited [52]. By using PBPK modeling, we offer model-informed dosing recommendations as a solution to a pressing clinical need—particularly valuable when traditional evidence is lacking.

  generalconclusions
  Keywords: similar model evidence dosing clinical based methodologies successfully implemented areas facing gaps instance optimizing gentamicin

Questions about this gap